A real scene: an association runs its annual conference, collecting sign-ups via a Google Form with manual bank-transfer reconciliation — 300 registrations, 40 payments that do not match, and a paper roll-call queue still running at start time. The organizers firefight all day and never get to greet the speakers. The value of an event site is not looks; it is money flowing in smoothly and people walking in smoothly. Here is the breakdown: when to self-host, how the process runs, real costs, and the most common traps.
When it fits vs. when it does not
Self-hosting an event site fits when:
- You run several events a year and want to build your own member/contact list (not leave it on a platform)
- Ticket prices are high and a 2–5% platform cut hurts
- Sign-up logic is complex: grouping, lottery, group orders, early-bird/bundles, identity checks
- You need registration data wired into an existing CRM/membership system
A ready-made platform is smarter when:
- One-off event, standard ticket types, fine with a service fee → a ticketing platform is fastest
- Very low budget, very tight timeline (live within a week)
- Free event, only collecting a list — a Google Form suffices
Alternatives matrix
| Option | Pros | Cons | Cost band |
|---|---|---|---|
| Ticketing platforms (e.g. KKTIX/Accupass) | Live same day, built-in traffic and check-in app | 2–5% service cut, list stays on platform, limited customization | No build, commission-based |
| Google Form + manual reconciliation | Free, zero learning | No payments, reconciliation hell, no check-in tool | NT$0 (but eats labor) |
| WordPress + ticketing plugin | Cheap, many plugins | Plugin dependency/maintenance burden, custom logic stalls | NT$30,000–80,000 |
| Custom event site (self-hosted) | Full control of logic and list, CRM-ready, cheapest long term | Higher upfront | From NT$80,000 |
Full process (with tools and deliverables)
- Week 1 — Planning: inventory ticket types, sign-up fields, payment methods, check-in flow. Deliverable: feature list + registration flowchart (tools: Figma/Notion).
- Weeks 2–3 — Sign-up and payments: registration form, ticket-type inventory, integrate ECPay/Stripe, auto-send confirmation emails and e-tickets (with QR). Deliverable: an orderable test environment.
- Week 4 — Admin and check-in: organizer admin (list export, refunds/changes, live sales numbers), QR check-in page (with offline backup). Deliverable: admin + check-in tool.
- Weeks 5–6 — Testing and rehearsal: payment load test, check-in flow rehearsal, go-live. Deliverable: acceptance checklist + production launch.
Full real-cost breakdown
- Development: standard NT$80,000–150,000; advanced (lottery/group orders/invoicing) from NT$150,000
- Payment fees (hidden, per transaction): credit card ~2.75–3.4%, straight off your margin — bake it into ticket price
- Notifications: email usually within free tiers; SMS ~NT$1–2 each, a real per-message cost for large events
- SSL (Let's Encrypt free), hosting: ~NT$6,000–20,000/year
- Maintenance: setup and testing hours before each event — best amortized with a maintenance contract
Implementation reality vs. client imagination
- Clients think the sign-up page takes two days; in reality payment integration + reconciliation + refund/change logic is the bulk of the hours — the form is the tip of the iceberg.
- Clients think QR check-in is always fast; in reality on-site network is the bottleneck, and without offline backup it jams on the day.
- Clients think it is over once tickets sell out; in reality refunds, name changes, invoices and post-event list remarketing are the real long tail.
Common traps and how to avoid them
- Trap: overselling (inventory not locked) → Fix: lock stock with a DB transaction at order time, auto-release on payment timeout.
- Trap: payment callback mis-wired, money paid but no ticket → Fix: treat server-to-server callback as source of truth, front-end redirect as backup only.
- Trap: on-site network dies, scans fail → Fix: check-in page supports offline list cache + a PDF backup export.
- Trap: personal data unencrypted, leaked after the event → Fix: encrypt sensitive fields, set a retention limit after the event.
- Trap: slow sign-up page loses half the conversions → Fix: watch LCP (Core Web Vitals), compress images, use a CDN.
Success metrics + 90-day roadmap
- Day 30: track sign-up conversion (visit → completed payment) and checkout-abandon rate; streamline checkout steps.
- Day 60: analyze source channels and ticket-type sales curves; adjust early-bird and marketing timing.
- Day 90: segment the event list for remarketing (next event's early bird); measure repeat-purchase and list monetization.
Decision checklist
- ☐ I run 4+ events a year, or a single event grosses 600k+
- ☐ I want to keep the registration list myself, not on a platform
- ☐ My ticket types/sign-up logic cannot be met by a ready-made platform
- ☐ The platform commission now exceeds the amortized cost of self-hosting
- ☐ I need registration data wired into an existing membership/CRM
- ☐ I can schedule a full check-in rehearsal before go-live
- ☐ I have baked payment fees into the ticket price
- ☐ I have written refund/change rules
- ☐ I have a backup plan for on-site network failure
- ☐ I will remarket to the list after the event
Six or more checked: self-hosting pays off for you. Fewer than four: run this event on a ready-made platform first.
FAQ
It is a one-time event — is a custom site worth it, or just use a ticketing platform?
One-off, standard ticket types, fine with a 2–5% service cut — a platform is fastest. But if you run several events a year, want to build a member list, sell high-priced tickets (where the cut hurts), or need custom logic (grouping, lottery, group orders), self-hosting pays off long term. The dividing line is roughly 600k+ annual revenue or 4+ events a year.
Should I use ECPay or Stripe for payments, and how much is the fee difference?
Depends on your audience. For local Taiwan payers needing convenience-store codes/ATM, ECPay is smoothest (credit card ~2.75–3.0%); for overseas cards or subscriptions, Stripe (~3.4% + a fixed fee). In practice many integrate both: local via ECPay, overseas via Stripe. Confirm rates on the official pages.
What goes wrong most on check-in day?
Three things: unstable on-site Wi-Fi jamming QR scans — always keep an offline backup list; peak-time crowding at entry — rehearse the flow and staffing; and no process for ticket changes/refunds, which causes scenes at the door. None are technical — they are about whether you rehearsed before go-live.
Roughly how much and how long for an event ticketing site?
Standard (sign-up form + payments + QR check-in + admin list): about NT$80,000–150,000, 4–6 weeks. Advanced (grouping/lottery, group orders, invoice integration, multi-session): from NT$150,000. Do not forget hidden costs: payment fees, SMS/email notifications, SSL and hosting annual fees.
Call to action
ScriptWalker builds event registration / ticketing websites (standard from NT$80,000), including payment integration, QR check-in and an organizer admin, and can wire into your existing membership/CRM. Want to confirm whether self-hosting pays off first? Book a free 30-minute consult and we will calculate this event's payment cost and timeline with you:
- Email: [email protected]
- Phone: 0916-224-047
- LINE: @ufv9089p